There is a lot of BS being disseminated about single CNN poll

[CajunBlazer]

No one bitched about it when it showed Hillary comfortably ahead.

It may be right, it may wrong. Sometimes the pollster pick a bad random sample, or a non-representative sample of people pick up the phone while other don't - it happens. It is also possible to blindly reach 100 times into a barrel filled with 1000 green marbles and 1000 red marbles and pick out a sample of 25 red marbles and 75 green marbles. Is it statistically likely? No, but it can and will happen occasionally.

What the media doesn't tell you is that the poll sample errors are written like this example where a candidate gets 50% of the responses: "This pole is expected to be accurate within plus or minus 3.5% 95% of the time" The media always leave that last part out because they believe most people are clueless about statistics and it would confuse them. What that statement means is 2.5% of the time the candidate's will actually get less than 46.5% (50% - 3.5%) of the vote and 2.5% of the time he/she will be more than 53.5% (50% + 3.5%). So polls can and sometimes will be "off". And yes, if the sample is not selected properly by the polling agency, they can be way off.

But here is the thing, most of the time polls from good companies are pretty accurate - that is they will within their sampling error most of the time. The CNN poll may be an outlier; but quit stressing over it - check other national polls taken about the same time period. A combination of poll results is better than the results of one poll.

However, whether we like it or not the national race is tightening significantly. However, we don't use national popular votes to pick our Presidents; it is the polls of swing states that are important. And guess what, polls of swing states are tightening significantly as well, but Hillary is still ahead in most of them, comfortably in some cases and not so comfortably in other very important ones.

Let that be a warning call - instead of fussing about the accuracy or lack of accuracy of polls, put your efforts into GETTING OUT THE VOTE.

[SaschaHM]

1. The thing about the CNN poll that is....

It's not that their sample is really bad, it's that the model that they then throw on Registered Voters (where Clinton is up +3) to make the likely voter numbers (where Trump is up) is bad. They're assuming a lot of things that weren't even the case in 2012 i.e. uneducated white voters being 50% of the electorate.

All polling companies do it, but CNN's model is clearly weighted with a specific electorate in mind and there is just next to little basis for that happening.

[Demsrule86]

2. The state polls matter

and with Georgia and Texas so close, it is clear she is ahead...the national polls are meaningless ...consider if everyone in a red red state say Idaho suddenly makes up their mind...it affects national polls but they were always going red anyway. The race may have tightened. I have my doubts however and do believe much is the media trying to make a comptetitive race...I believe we may actually see a landslide when all is said and done.

[CajunBlazer]

4. Maybe but Florida, Ohio and North Carolina are too close for confort

A combination all recent polls on Nate Silver's website show Hillary is ahead by 2.5% or less in all three states which cumulatively have 62 electoral votes and the race in all swing states, including those three, have been continually tightening since August 14th. If the current trends continue, they will tighten even more. Those three state are the only ones which giving Hillary a comfortable lead in the electoral college figures. Without them she would have 278 votes - much to close for comfort.

[Demsrule86]

13. North Carolina?

Are you kidding? We almost never get North Carolina...we got it in 08 ...as for the other two...Hillary is still ahead and internal polls show her pulling ahead in Ohio...don't know about Florida but I doubt Trump will win ... the minority vote will turn out...I would remind you though we have many different ways to win these days...Obama could have lost both Ohio and Florida and still won. The GOP has only the Bush model...and many believe Bush did not win either of those elections...and it was before the GOP lost Virginia...The GOP is in a bad place electorlly speaking.

[CajunBlazer]

15. I strongly suggest that you read the article the OP that I about to publish

The article in question is:

Election 2016 – Too Close for Comfort

[Demsrule86]

3. It wasn't the same poll also...they had 50% non-college and more GOP types

bad sample=bad poll.

[CajunBlazer]

5. Not necessaryily a bad poll

If it were real bad the other national polls would be much different; Here are the most recent national polls.

CNN/ORC Clinton 43, Trump 45, Johnson 7, Stein 2 Trump +2

LA Times/USC Clinton 44, Trump 44 Tie

NBC News/SM Clinton 41, Trump 37, Johnson 12, Stein 4 Clinton +4

IBD/TIPP Clinton 39, Trump 39, Johnson 12, Stein 3 Tie

The average: Clinton +0.5%

It's time to face the music - the national race has tightened significantly.

But you should be stressing over the swing state polls which have which have been tightening since the convention.

[factfinder_77]

6. IBD/TIPP : HRC only gets 29 % of male vote and 67 % of black/hispanic vote.

And the LA Times/USC has been on the front page of drudge the last month...has been the only one Trump has been leading Clinton (until now that is)

[CajunBlazer]

9. Quit complaining about the polls and GOTV

[Lord Magus]

7. The LA Times poll has been infamously bad throughout the entire campaign.

When half of the polls used for an "average" are blatantly garbage, that makes the average also garbage.

[CajunBlazer]

8. The LA Times poll is not a bad poll!

The LA Times polls compensates for post convention bounces which traditionally do not last. Therefore the poll minimized Hillary's lead for a period of time after the convention. Had her bounce after the convention had lasted, you would have seen her numbers in that poll increase by now. But guess what, the bounce has not lasted like we would have liked it to, so the poll was correct to discount it.

Again, it is time to quit complaining about polls we don't like and GOTV.

[Lord Magus]

11. Yes, it absolutely is a bad poll.

It's been massively pro-Trump since before the conventions. Its nonsensical methodology has been ripped to shreds all year. Quit trying to claim it's something worth paying attention to, that's on the same level as saying Fox's reporting is worthwhile.

[Demsrule86]

12. It's a bad poll

When traditionally red states are close or even tied...yeah it is a bad poll.

[Demsrule86]

14. National polls are meaningless at this point...

As people make up their mind in red states...the national polls will shift but the Dems were never going to carry those states anyway. Look at the state polls. The fact that traditionally red states are close does not bode well for the orange cheetoh.

[Lebam in LA]

10. It is 1 poll out of dozens

When the majority of polls show him winning, I'll worry